SPI Map Interpretation

The SPI index was developed by T.B. McKee, N.J. Doesken, and J. Kleist in 1993. The data and climate division boundaries are from the National Climatic Data Center. The Western Regional Climate Center uses the NCDC data to calculate SPI values for each climate division. The information is then reclassed and mapped at the National Drought Mitigation Center using a Geographic Information System.

A one-month SPI map is very similar to a map displaying the percent of normal precipitation for a month. It is actually a more accurate representation of monthly precipitation because the distribution has been normalized (see detailed discussion of the Percent of Normal). Because the 1-month SPI reflects relatively short-term conditions, its application can be related closely with short-term soil moisture and crop stress, especially during the growing season. The 1-month SPI may approximate conditions represented by the Crop Moisture Index (CMI).

Interpretation of the 1-month SPI may be misleading unless climatology is understood. In regions where rainfall is normally low during a month, large negative or positive SPIs may result even though the departure from the mean is relatively small. Several specific examples demonstrate this limitation.

Examine the 1-month SPI for June 1995 as a first example. The South Coast Drainage Climate Division in California has an SPI of +3.11, giving the impression that the month of June was very wet. Indeed, the very large positive SPI means that the amount of rain that did fall during that June was greater than the third standard deviation and that this occurs approximately once in 100 years. However, the normal June precipitation for this climate division is 2.5 mm (0.10 in.) (see Los Angeles climograph), and the actual precipitation total for June 1995 was 15.2 mm (0.60 in.).Thus, it was highly unusual but hardly a flooding event. In comparison, the Willamette Valley Climate Division in Oregon had a 1-month SPI in February 1996 of +1.97. In this climate division, 371.9 mm (14.64 in.) of precipitation fell in February, which was 211.6 mm (8.33 in.) above the normal total for the month. Major flooding occurred along the Willamette and Columbia Rivers as a result of these high rainfall amounts. This comparison emphasizes the necessity of understanding the climatology of an area when examining the 1-month SPI.

A second example of how the 1-month SPI may be misleading is demonstrated by the 1-month January 1996 SPI for the Central Climate Division in Nebraska. Normally, the North American Great Plains get very little precipitation during the mid-winter months (see the Lincoln, Nebraska, Lubbock, Texas, and Bismarck, North Dakota climographs), and the January normal for this climate division is only 10.4 mm (0.41 in.). In 1996, the January precipitation total was 24.9 mm (0.98 in.). This is 239% of normal, and it gives an SPI of +1.43. However unusual the January precipitation totals were for this climate division, as well as others in southeastern Nebraska, these totals still provided a relatively small amount of moisture and did little to diminish the longer-term dry period in the region (see 9-month SPI for the end of February 1996).

The 1-month SPI can also be misleading with precipitation values less than the normal in regions with a small normal precipitation total for a month. This can be seen in the 1-month SPI for the end of February 1996 for the Southeastern Plains Climate Division in New Mexico. The SPI is -1.76. This large negative SPI gives the impression that February was very dry in this climate division. It was indeed very dry (zero precipitation fell during the month), but February is normally a dry month in this location (see climograph for Lubbock, Texas), so the departure from normal was only 10.4 mm (0.41 in.).

These are only a few of many possible examples. As with a percent of normal map, useful information is contained in the 1-month SPI maps, but caution must be observed when analyzing these maps.

The three-month SPI provides a comparison of the precipitation over a specific 3-month period with the precipitation totals from the same 3-month period for all the years included in the historical record. In other words, a 3-month SPI at the end of February compares the December–January–February precipitation total in that particular year with the December–February precipitation totals of all the years.

A 3-month SPI reflects short- and medium-term moisture conditions and provides a seasonal estimation of precipitation. In primary agricultural regions, a 3-month SPI might be more applicable in highlighting available moisture conditions than the slow-responding Palmer Index. A 3-month SPI at the end of August in the U.S. Corn Belt would capture precipitation trends during the important reproductive and early grain-filling stages for both corn and soybeans. Meanwhile, the 3-month SPI at the end of May gives an indication of soil moisture conditions as the growing season begins.

It is important to compare the 3-month SPI with longer time scales. A relatively normal 3-month period could occur in the middle of a longer-term drought that would only be visible at longer time scales. A good example of this can be illustrated by the four Climate Divisions in southeastern Nebraska. The 3-month SPI for the end of March 1996 showed these Climate Divisions in the “near normal” category, largely because of above-normal precipitation in January 1996.

However, dry conditions across this region began in June 1995, and can be seen in the 6-month SPI for the end of March 1996. Looking at longer time scales, as in this example, would prevent a misinterpretation that any “drought” might be over.

As with the 1-month SPI, the 3-month SPI may be misleading in regions where it is normally dry during that 3-month period. Large negative or positive SPIs may be associated with precipitation totals not very different from the mean. This caution can be demonstrated with the Mediterranean climate of California, where very little rain falls from April to September (see climographs for Los Angeles, San Francisco, and Fresno). Thus, a 3-month SPI for the end of August in any Climate Division in California will compare the historic precipitation totals for June–July–August. Because this is a time period with little rain, these historic totals will be small and relatively small deviations on either side of the mean could have large negative or positive SPIs.

The 6-month SPI compares the precipitation for that period with the same 6-month period over the historical record. For example, a 6-month SPI at the end of September compares the precipitation total for the April–September period with all the past totals for that same period.

The 6-month SPI indicates medium-term trends in precipitation and is still considered to be more sensitive to conditions at this scale than the Palmer Index. A 6-month SPI can be very effective in showing the precipitation over distinct seasons. For example, a 6-month SPI at the end of March would give a very good indication of the amount of precipitation that has fallen in California during the very important period from October through March (see 6-month SPI for the end of March 1996). Information from a 6-month SPI may also begin to be associated with anomalous streamflows and reservoir levels.

The 9-month SPI provides an indication of precipitation patterns over a medium time scale. Droughts usually take a season or more to develop. SPI values below -1.5 for these time scales are usually a good indication that fairly significant impacts are occurring in agriculture and may be showing up in other sectors as well. For example, the 9-month SPI map for the end of May 1996 shows that many climate divisions from Utah and Arizona to Texas have SPI values less than -1.5. As a result, President Clinton authorized $70 million in aid for the drought-stricken states of Texas, Oklahoma, Kansas, New Mexico, and Colorado.

Some regions of the country may find that the pattern displayed by the map of the Palmer Index closely relates to the 9-month SPI maps. For other areas, the Palmer Index is more closely related to the 12-month SPI. The Palmer Index maps are updated each week, although the patterns usually do not change significantly. The SPI maps are updated at the end of each month.

The SPI at these time scales reflects long-term precipitation patterns. A 12-month SPI is a comparison of the precipitation for 12 consecutive months with the same 12 consecutive months during all the previous years of available data. Because these time scales are the cumulative result of shorter periods that may be above or below normal, the longer SPIs tend toward zero unless a specific trend is taking place. The 12-month SPI for the end of February 1996 compared the precipitation totals for the March 1995–February 1996 period with similar periods in history. Most of the climate divisions across the United States were in the “near normal” category. The large positive SPIs across South Dakota indicate the very wet conditions that existed for this 12-month period. Indeed, 1995 was the wettest year on record for the entire state of South Dakota, and this is still shown in the 12-month SPIs for the state.

SPIs of these time scales are probably tied to streamflows, reservoir levels, and even groundwater levels at the longer time scales. In some locations of the country, the 12-month SPI is most closely related with the Palmer Index, and the two indices should reflect similar conditions.